Send diagnostic information to Caplin Support

This page describes how to send diagnostics, logs, and configuration files for a Caplin component to Caplin Support.

Requirements

To collate diagnostics for a Caplin component, you require:

  • Caplin Platform Diagnostics. A set of open-source scripts provided by Caplin that collates diagnostic information on a running component or a core file.

  • GNU Debugger (gdb package) installed on the component’s host. Required by Caplin Platform Diagnostics to generate and analyse core files.

    $ sudo yum install gdb

Collating diagnostics for a crashed component

Follow the steps below to send Caplin Support diagnostics for a crashed component:

  1. If the crashed component dumped a core file, run the Caplin Platform Diagnostic’s caplin-core-diagnostics.sh script on the component’s host.

    Syntax
    caplin-core-diagnostics.sh core [binary]
    Example
    $ ./caplin-core-diagnostics.sh ~/dfw1/servers/Liberator/core.3581

    For full instructions and requirements, see the Caplin Platform Diagnostics README.

  2. Create a single tar archive containing whichever of the diagnostic files below are present:

    Caplin Platform Diagnostics output

    diagnostics-<host>-<core_file>-<timestamp>.tar.gz

    JVM HotSpot error file

    component_working_dir/hs_err_pid<pid>.log

    JVM heap dump

    component_working_dir/var/java_pid<pid>.hprof

    JVM GC logs

    component_working_dir/var/gc.log.*

    Log files for the incident period

    component_working_dir/var/*

    Caplin configuration files

    <dfw_root>/global_config/*.conf
    <dfw_root>/global_config/overrides

  3. Log in to Caplin’s secure File Upload Facility and upload the archive to Caplin Support.

Collating diagnostics for a running component

Follow the steps below to send Caplin Support diagnostics for a running Caplion component (without terminating the component’s process):

  1. Run the Caplin Platform Diagnostic’s caplin-process-diagnostics.sh script on the component’s host.

    Syntax
    caplin-process-diagnostics.sh pid
    Example
    $ ./caplin-process-diagnostics.sh 3581

    For full instructions and requirements, see the Caplin Platform Diagnostics README.

  2. Create a single tar archive containing whichever of the diagnostic files below are present:

    Caplin Platform Diagnostics output

    diagnostics-<host>-<binary>-<pid>-<timestamp>.tar.gz

    JVM GC logs

    component_working_dir/var/gc.log.*

    Log files for the incident period

    component_working_dir/var/*

    Caplin configuration files

    <dfw_root>/global_config/*.conf
    <dfw_root>/global_config/overrides

  3. Log in to Caplin’s secure File Upload Facility and upload the archive to Caplin Support.

Working directory locations

The location of a component’s working directory depends on whether it was deployed within or without a Caplin Deployment Framework.

Components deployed within a Caplin Deployment Framework
Component Working Directory

Liberator

<dfw_root>/servers/Liberator/

Transformer

<dfw_root>/servers/Transformer/

Integration adapter

<dfw_root>/kits/<blade_name>/<version>/DataSource/

Components deployed without a Caplin Deployment Framework
Component Control script Working Directory

Liberator

<deploy_root>/etc/rttpd

<deploy_root>

Transformer

<deploy_root>/etc/transformer

<deploy_root>

UPAsrc adapter

<deploy_root>/etc/upasrc

<deploy_root>